Home / Position Sizing / Fixed-Fractional Position Sizing
SizingIntermediate

Fixed-Fractional Position Sizing

Fixed-fractional position sizing risks a constant fraction of current account equity on every trade, so the rupee bet grows as the account grows and shrinks as it falls, producing geometric compounding and a self-limiting response to drawdowns.

Quick answer: Fixed-fractional position sizing risks a constant fraction of current account equity on every trade, so the rupee bet grows as the account grows and shrinks as it falls, producing geometric compounding and a self-limiting response to drawdowns.

In simple words

Fixed-fractional sizing means you risk the same percentage of your current capital on each trade, for example 1 percent, rather than a fixed number of lots. When the account grows you bet more rupees; when it shrinks you bet less. This makes the account compound in good times and automatically pull back exposure after losses, which is why it is one of the most widely used sizing rules.

Purpose

Fixed-fractional sizing exists to make risk scale with capital rather than stay static, giving geometric growth while building in an automatic de-risking response to losing streaks so that no single run of losses is fatal.

Visual explanation

Fixed-Fractional Position Sizing

The rupee bet scaling up and down in proportion to current account equity as a constant fraction is risked each trade.

Risk-Based Position SizingCapital×Risk %Stop distance×Point value=Quantityround down to lot sizerisk a fixed fraction of capital per trade

Professional explanation

The core mechanism and formula

Fixed-fractional sizing sets the quantity so that a fixed fraction f of current equity is at risk if the stop is hit. The number of units is the rupee risk budget divided by the per-unit risk: units = (f × capital) ÷ (stop distance × point value). Because capital is re-read before every trade, the bet compounds: a rising account raises the rupee bet, a falling account cuts it. This is the defining property that separates it from fixed sizing, where the quantity ignores the account balance entirely.

Why it compounds geometrically

When each trade risks the same percentage, gains and losses multiply rather than add. A sequence of returns r1, r2, r3 applied to capital produces a terminal wealth proportional to the product of (1 plus each return), not their sum. Over a long sample with a positive edge this geometric compounding dramatically outgrows the additive path of fixed sizing. The same mathematics, however, means the growth rate is governed by the geometric mean of outcomes, which is always less than the arithmetic mean, so volatility drags on compounded growth in a way fixed sizing hides.

The self-limiting response to drawdowns

Because the rupee bet is a fraction of shrinking equity, fixed-fractional sizing automatically reduces exposure during a losing streak. After a 20 percent drawdown the account is smaller, so the next 1 percent bet is 1 percent of a smaller number, cutting the absolute risk. This anti-martingale behaviour is the reason risk of ruin under fixed-fractional sizing is far lower than under an equally aggressive fixed scheme: you cannot reach zero by linear steps because each loss shrinks the next bet. The cost is a slower recovery, since you also bet less on the way back up.

Choosing the fraction and the drawdown trade-off

The fraction f directly controls the trade-off between growth and drawdown. Too small and the account barely moves; too large and normal variance produces punishing equity swings, because percentage drawdowns scale roughly with f. There is an optimal-growth fraction, given by the Kelly criterion, beyond which increasing f actually lowers long-run growth while still raising volatility. Practitioners almost always bet a fraction well below the Kelly optimum, often a half or a quarter of it, trading some growth for a much smoother ride and a wider margin against estimation error.

Backtesting implications and path dependence

A fixed-fractional backtest is path dependent: the same set of trades in a different order produces a different terminal wealth and a different maximum drawdown, because the bet size at each point depends on the running equity. This means a single compounded equity curve is only one realisation, and Monte Carlo reshuffling of the trade sequence is essential to understand the distribution of outcomes and the realistic drawdown range. It also means you must feed the sizing model the true per-trade risk, including realistic stop distances after slippage and costs, or the compounded figures will be optimistic.

Formula

units = (f × capital) ÷ (stop distance × point value)

units = number of shares, lots or contracts to trade (round down to a whole lot in F&O); f = fraction of capital risked per trade, e.g. 0.01 for 1 percent; capital = current account equity in rupees; stop distance = points between entry and stop-loss; point value = rupee change per one-point move for one unit (for a Nifty lot of 75, point value = Rs 75 per index point).

Fixed-fractional vs Fixed size

AspectFixed-fractionalFixed size
Held constantFraction of capital riskedQuantity traded
GrowthGeometric, compoundingAdditive, near-linear
After lossesBet shrinks automaticallyFraction at risk rises
Risk of ruinMuch lower for equal aggressionHigher, linear path to zero
Path dependenceStrong; order of trades mattersWeaker; each trade independent

Practical example

Illustrative example (Indian market)

On capital of Rs 5,00,000 you risk f = 1 percent per trade, so the risk budget is Rs 5,000. Your Bank Nifty setup has a stop 120 points away and the point value is Rs 15 per point per lot of 15 units, so per-lot risk is 120 × 15 = Rs 1,800. Units = 5,000 ÷ 1,800 = 2.78, which you round down to 2 lots. After a good run the account reaches Rs 7,00,000, so the same 1 percent is now Rs 7,000 and, with the same stop, you would size 7,000 ÷ 1,800 = 3.8, rounded to 3 lots. The bet grew with the account without you changing the rule, which is the compounding effect fixed-fractional sizing is built to deliver.

Because NSE lots are indivisible, the rounded-down unit count means a small account crosses sizing thresholds in jumps. On Rs 5,00,000 at 1 percent you might afford only 2 Nifty lots, and you will not reach 3 lots until equity and the risk budget grow enough to clear the whole-lot boundary, so the compounding is stair-stepped rather than smooth for retail F&O traders.

Limitations

  • Terminal wealth is path dependent, so one compounded curve is only a single realisation of many possible orders
  • The geometric mean governs growth, so volatility drags on compounded returns more than fixed sizing reveals
  • Recovery from drawdown is slow because the bet also shrinks on the way back up
  • Lot indivisibility forces rounding, so a small F&O account jumps between sizes rather than scaling smoothly
  • It relies on an accurate stop distance and per-trade risk; underestimating them makes the compounded curve optimistic

Why it matters in practice

  • It converts a per-trade edge into geometric account growth while automatically de-risking after losses
  • It makes drawdown depth a direct function of the chosen fraction, which is the main lever a trader controls

Common mistakes

  • Betting at or above the Kelly-optimal fraction, which raises volatility while lowering long-run growth
  • Reading capital once at the start instead of before each trade, which turns it back into fixed sizing
  • Ignoring path dependence and trusting a single compounded curve instead of Monte Carlo reshuffling
  • Forgetting to round down to whole lots, overstating the achievable size on a small F&O account
  • Using an unrealistically tight stop in the formula, which inflates unit count and understates risk
  • Confusing fraction of capital risked with fraction of capital deployed as notional exposure

Professional usage

Institutional and professional systematic traders treat the risked fraction as the central risk dial and set it deliberately below the Kelly optimum, commonly at a half or quarter Kelly, to buy robustness against estimation error and fat tails. They validate the sizing with Monte Carlo reshuffling of the trade sequence to see the distribution of terminal wealth and drawdown rather than a single path, and they feed the formula post-cost, post-slippage stop distances so the compounded figures are honest. The fraction is often reduced further during regime uncertainty or after a strategy has underperformed its expected drawdown envelope.

Key takeaways

  • Fixed-fractional sizing risks a constant fraction of current capital, so bets compound with the account
  • Units = (f × capital) ÷ (stop distance × point value), re-read before every trade
  • It de-risks automatically after losses, giving a far lower risk of ruin than aggressive fixed sizing
  • Percentage drawdown scales with the fraction f, which is the main lever you control
  • Results are path dependent, so validate with Monte Carlo rather than one compounded curve

Frequently asked questions

What is fixed-fractional position sizing?
It is a rule that risks a constant fraction of your current account equity on each trade, so the rupee bet grows as the account grows and shrinks as it falls. This produces geometric compounding and an automatic reduction of exposure during losing streaks.
What is the fixed-fractional sizing formula?
Units = (f × capital) ÷ (stop distance × point value), where f is the fraction risked such as 0.01, capital is current equity, stop distance is the points to your stop, and point value is the rupees per point for one unit. Round the result down to whole lots in F&O.
How is fixed-fractional different from fixed sizing?
Fixed sizing holds the quantity constant while risk drifts, whereas fixed-fractional holds the risked percentage constant so the rupee bet moves with the account. The first grows additively; the second compounds geometrically.
Why does fixed-fractional sizing compound?
Because each trade risks the same percentage, returns multiply rather than add. Terminal wealth is proportional to the product of one plus each return, so a positive edge grows the account geometrically over a long sample.
How does it reduce risk of ruin?
Since the bet is a fraction of shrinking equity, each loss automatically cuts the next bet, so you approach zero asymptotically rather than in fixed linear steps. This anti-martingale behaviour makes ruin far less likely than an equally aggressive fixed scheme.
What fraction should I risk per trade?
The fraction is a growth-versus-drawdown choice; percentage drawdown scales roughly with it. Most practitioners bet well below the Kelly-optimal fraction, often a half or a quarter of it, to gain robustness against estimation error and fat tails.
Does a bigger fraction always mean more growth?
No. Beyond the Kelly-optimal fraction, increasing the bet lowers long-run growth while still raising volatility, because compounded growth follows the geometric mean. Over-betting can turn a positive edge into a shrinking account.
Why are fixed-fractional results path dependent?
Because the bet size at each trade depends on running equity, the same trades in a different order give a different terminal wealth and drawdown. A single compounded curve is only one realisation, which is why Monte Carlo reshuffling is important.
How does lot indivisibility affect fixed-fractional sizing?
In NSE F&O you must round down to whole lots, so a small account crosses size thresholds in jumps rather than scaling smoothly. The compounding is therefore stair-stepped for retail traders.
What inputs does the formula need?
It needs the risked fraction, the current capital, the stop distance in points, and the point value per unit. For a Nifty lot of 75 the point value is Rs 75 per index point, so a 100-point stop risks Rs 7,500 per lot.
Why does recovery from drawdown feel slow?
Because the same rule that shrinks your bet after losses also keeps the bet small while the account recovers, so you rebuild more slowly than a fixed scheme would. Lower drawdown risk is bought at the cost of slower recovery.
Should I use the arithmetic or geometric mean to judge it?
Judge compounded growth by the geometric mean of outcomes, which is always at or below the arithmetic mean. The gap between them widens with volatility, which is why fixed-fractional growth is dragged down by variance.
Can I convert a fixed-size backtest to fixed-fractional?
You must re-run the identical signals through the sizing formula rather than rescale the result, because the compounded path and drawdowns change once the bet size depends on running equity.
Is fixed-fractional the same as risk-based sizing?
They are closely related. Risk-based sizing sets quantity from a money or percentage risk and a stop distance; fixed-fractional is the special case where that risk is expressed as a constant fraction of current equity on every trade.

Voice search & related questions

Natural-language questions people ask about Fixed-Fractional Position Sizing.

What is fixed-fractional position sizing in plain words?
It means risking the same percentage of your account on every trade, like 1 percent, so you bet more rupees when the account grows and less when it shrinks. That makes it compound in good times and pull back after losses.
How do I calculate fixed-fractional size?
Take the fraction times your capital to get your rupee risk, then divide by the stop distance times the point value. That gives the number of lots, which you round down in F&O.
Why does risking a fixed percentage protect me?
Because after a loss your account is smaller, so the same percentage is a smaller bet. You keep shrinking the bet as you lose, so it is very hard to go all the way to zero.
What percentage should I risk per trade?
Most traders stay well under the mathematically optimal amount, often around a half or a quarter of it, because betting the full optimum is a very rough ride and depends on estimates that are never exact.
Does fixed-fractional grow faster than fixed sizing?
Over a long run with a real edge, yes, because it compounds. But it also has deeper percentage drawdowns and recovers more slowly, so the smoother growth comes with bigger swings along the way.
Why do I have to re-check my capital before every trade?
Because the whole point is that the bet follows the account. If you read your capital once and never update it, you are back to fixed sizing and lose the automatic scaling.

Sources & references

    Last reviewed 12 July 2026. Educational content only — not investment advice. Markets and rules change; verify current conventions with SEBI, NSE/BSE and your broker.

    ← Previous
    Fixed Position Sizing
    Educational content only — not investment advice. Examples use illustrative numbers and simplified models. Backtested results are hypothetical and trading derivatives involves substantial risk. See our Risk Disclosure and SEBI Disclaimer.